So I guess "A.I." is the new buzzword for how we describe all digital technology going forward. FWIW I thought this was a better article that described the actual tech used by KoBold, https://spectrum.ieee.org/ai-mining. Now, that article was written by the KoBold CEO, so there are certainly parts of it I'd take with a giant chunk of salt, but I think it's easier to read that article (and read past some of the AI buzzwords) to see how they're probably using (a) better surveying tech and (b) standard machine learning techniques to generate maps of potential deposits.
> So I guess "A.I." is the new buzzword for how we describe all digital technology going forward
I wonder if any companies are getting deals on compute for making a big splashy deal out of the part ML played in these processes. Kind of a B2B meets Twikstogrube Influencer marketing strategy, but instead of companies having cachet because a bunch of social media followers find them appealing, they actually manifest things in the physical world. That is a big hole in the Generative AI company sales pitch for a) non-early-adopter potential customers, and b) many others looking uneasily at the kind of resources they're tearing through when the only tangible things they've seen from it are a pitches for features they never asked for and don't care about, and very concerning faked images and videos for extortion, bullying, porn, and political shit. I'm not saying those things are all its good for, but the communication about the real-world value of this stuff has been pretty lacking, and the drawbacks have been understandably shouted from the rooftops, so they're probably preeeetttyyy thirsty for stories like this.
“That means the conventional predictions are largely inference—and worse, they result in unquantified uncertainty.”
Wild claim given the fact that Gaussian process regression / Kriging was invented in the 1960s in geoscience to do exactly what the article claims only their models do: “quantify uncertainty, which in turn guides our data collection, as the most uncertain rocks often represent the most valuable ones to sample”
IMO AI means neural net. I get that people use it for other things, but that's what I use it to mean - there is just no other term that's easy to say. And at this point the idea of breaking problems down into "neurons" and activation patterns is inherent to most AI models. Here though the keywords are "ensemble machine learning" and "Bayesian" - they could have used a neural net for the machine learning but most likely it is just XGBoost or similar. https://ia.acs.org.au/article/2021/the-ml-technology-looking... mentions they are also doing full-physics joint inversions and computer vision, perhaps the vision is a neural net.
I dug into this a tiny but and learned that the ACM started having conferences in the early 1950s - https://dl.acm.org/doi/pdf/10.1145/30408.30418 - while the famous Dartmouth workshop that coined the term AI was later, in 1956. So I agree, it looks like computer science was already establishing itself as a discipline before the term AI started to be used.
There's been some overlap naturally, e.g. Yershov or Hopper attending both types events early on. But hardly any of content in say TAoCP was considered AI at any point.
my friend, XGBoost is not a neural net-based method, it represents a best-of-breed for an alternative family of methods.
"While the XGBoost model often achieves higher accuracy than a single decision tree, it sacrifices the intrinsic interpretability of decision trees. For example, following the path that a decision tree takes to make its decision is trivial and self-explained, but following the paths of hundreds or thousands of trees is much harder."
Unless the parent edited their comment, you misread it, because they are specifically saying XGBoost is not a neural network:
> Here though the keywords are "ensemble machine learning" and "Bayesian" - they could have used a neural net for the machine learning but most likely it is just XGBoost or similar.
I.e. they could have used a neural network but they probably used something else, that something else being XGBoost.
Copper mining is insane - vast vast machines (like apartment buildings on wheels) because the ore extraction rate is so low.
But this is a boon to a democratic light in the Southern Africa, (https://georgetownsecuritystudiesreview.org/2023/05/05/zambi...) here’s hoping they negotiate hard with the miners and stuff as much revenue into solid projects (roads hospitals schools etc)
The resource curse, also known as the paradox of plenty or the poverty paradox, is the phenomenon of countries with an abundance of natural resources (such as fossil fuels and certain minerals) having less economic growth, less democracy, or worse development outcomes than countries with fewer natural resources. There are many theories and much academic debate about the reasons for and exceptions to the adverse outcomes. Most experts believe the resource curse is not universal or inevitable but affects certain types of countries or regions under certain conditions.
I take your resource curse and raise you a Prebisch-Singer hypothesis
> Prebisch–Singer hypothesis argues that the price of primary commodities declines relative to the price of manufactured goods over the long term, which causes the terms of trade of primary-product-based economies to deteriorate. As of 2013, recent statistical studies have given support for the idea.
Derived demand (copper) is likely to be more price sensitive than demand for the end good (AI services) over the long run due to substitution that can occur in factor inputs. Meaning Prebisch-Singer's true again.
I imagine that assumes we can find large new deposits to meet increasing demand. That's not happening with copper. The deposit in this article is an exception.
It doesn't assume. Hence 'substitute'. long run refers to the amount of time for prices to adapt, not a short run dependence on one thing.
Hence change in factor inputs.
An example of a change in factor inputs was, for example, whale oil which was a big deal for lighting systems in the 19th century. But as cheap whales were running out, so whale oil, thought to be so important for many and which there were technical improvements in sourcing and usage, was shifted away from.
As will be the case for copper, as it's a derived demand. The derived demand will be more price sensitive than the demand for the end service (lighting in the case of whales, or Open AI for copper). Not happening with copper in the short run will only make the long run change happen even (as time passes) faster.
Ooohhh - intriguing.
And presumably lines up with the Prebish-Singer hypothesis- that ports can switch to substitute goods - also interesting that almost every major city globally is a port city / river city
A truly dense population can not be sustained in a location without local resources unless there is a way to import at least the necessities for survival from somewhere else where they are more abundant.
Some of the best places for that, naturally are port cities, and for millennia merchant marine moves more goods more sensibly than most.
Inland cities can get big more easily when there are abundant local resources, well developed, and if there is some huge excess of something like timber, coal, gold, or whatever that is widely desirable. A local market will develop first in the land of abundance. Then if it's possible to arbitrage in a world market, the cost will be paid to shuttle commodities to an international port. Where the trading merchants will be able to buy low and sell high in a way that is further out of reach for the local producers.
Eventually the traders make more money on the same tonnage of natural resource flow than the extractors do, but the extractors got there first.
Depending on what the resource owners do with that early advantage, and whether there is a regime in place which values the local resource more than human life, and stuff like that has a much bigger effect on the imbalance between producer and trader, as well as local versus international prosperity, and that's with raw commodities not yet subject to value-added manufacturing.
Seems to me port cities mostly arose due to upstream export needs before they were utilized as major import hubs and mercantile centers.
I say it's the cheapness and abundance of commodities that makes it easier to add value for quite a long chain, and when you get into manufacturing it can be a big jump.
Services, technology and education are all worth money on their own but it does seem to take quite a bit of resource abundance at some point for somebody to be able to pay the bills.
Especially when resources were developed generations ago, there can be so many layers between the commodity and the consumer that things like service, tech, and education can be valued in their affect on each other without any direct correlation to a particular commodity or manufactured product.
I would argue Australia suffers from a lite version of the resource curse. There's undue control over politicians and resulting political resistance to invest in things that would diversify economic complexity or go against mining interests. Norway however is a strong counter example.
The key is whether or not a country, it's people and their representatives, are in control of the deal making wrt the resources within their boundaries.
Australia and Canada are, their indigenous people less so, and we can argue about the quality of the many resource deals within Australian and Canadian borders - overall they do less well than Norway.
This is in strong contrast to many African countries, Papua, and elswhere about the globe where often the key parts of government are wholly in the pocket of outside transnational corps who frequently have small divisions of PMC's (private militay contractors) for 'security' and land deals are forced through with near zero compensation to former land holders and NSR (Net Smelter Returns) | leasing returns to the country and people are near non existant.
The reality of what another peer commenter in this thread decsribed as
> But this is a boon to a democratic light in ... Africa
is anything but. eg: US PMC's in Africa .. acting for multiple clients, including China .. but not for Africans.
> Australia and Canada are [in control of the dealmaking wrt resources]
It's not a binary. It's a spectrum. The capability of Australians to control the deal making is diminished by the control that the mining industry has over elected representatives.
Please don't incorrectly paraphrase | strawman my comments.
Your point is implicit within:
> and we can argue about the quality of the many resource deals within Australian and Canadian borders - overall they do less well than Norway.
"The mining industry" should include energy extractors who, IMHO, do more harm to Austrlia than mining - many of the mining operations (not all by any means) are majority Australian owned|controlled with that money staying within Australia (even if with individuals rather than spread out across the entire community).
Even significant energy extractors such as Santos are Australian companies .. it's literally an acronym of South Australia Northern Territory Oil Search, but they're no angels although perhaps arguably better than the non-Australian gas operators.
Historically speaking, the people and the environment come a very, very distant second and third place to mine owners.
Even moreso these days because China and North Korea have sponsored so much infrastructure in Africa over 20 years as European colonisation has pulled out, and bought so many mines and in-field processing plants that the local people are just an ablative grease for mechanised and chemical mining.
> We need 25 times as much cobalt as we currently mine
Is that really true? I've lost significant money investing in explorers that hoped the Cobalt price would stay high. It hasn't, but also its hard to compete with Congo's mines.
Well... most current lithium battery chemistries absolutely require cobalt, and given the demand for EVs and grid-scale PV storage there's bound to be a lot of interest for it.
Unfortunately for cobalt miners, there's significant r&d investment into chemistries free of materials which are dominantly sourced from questionable countries/conditions - some driven by preparation for trade conflicts, and some driven by law (EU supply chain / anti slavery acts). And that is yielding its first results, e.g. [1].
Not yet, NMC Li-ion is still the largest represented in grid scale battery storage [1]. Pumped hydro is dominant overall in storage. LFP represents an increasing share of new grid scale battery storage, though [2]
When I was last in Zambia, I saw truck after truck full of huge copper sheets lined up at the border waiting to pay duty. I don't think it is any big news to anybody that there's copper there. It sounds like the AI just help find the extents of this particular deposit, not that it surprised anybody with secret copper.
My imagination tells me that they 'fed' it with all geographical/geological data available, and cross referenced them with various locations of mines (different ores)(gold, cobalt, copper, etc.) and the "machine" "figured out" that (silly example) river + mountain + over500m + earthquakes (or lack of) + various other parameters = so-and-so ore.
There may be some 'coincidences'/similarities that require many parameters that the eye misses, while the "AI" can combine far more parameters.
The thought experiment makes no sense. Once the basilisk is created it has no reason to create an incentive structure that applies to the past because the past is fixed. If the basilisk is actually superintelligence it's not going to fall for the sunk cost fallacy.
